Hydrophosphorylation of Alkynes Catalyzed by ... - ACS Publications

Feb 8, 2018 - Chang-Qiu Zhao,. ‡,§ and Li-Biao Han*,§. †. College of Material and Chemical Engineering, Hainan University, Haikou, Hainan 570228...
0 downloads 0 Views 4MB Size
Article pubs.acs.org/JACS

Cite This: J. Am. Chem. Soc. 2018, 140, 3139−3155

Hydrophosphorylation of Alkynes Catalyzed by Palladium: Generality and Mechanism Tieqiao Chen,†,§ Chang-Qiu Zhao,‡,§ and Li-Biao Han*,§ †

College of Material and Chemical Engineering, Hainan University, Haikou, Hainan 570228, China College of Chemistry and Chemical Engineering, Liaocheng University, Liaocheng, Shandong 252059, China § National Institute of Advanced Industrial Science and Technology (AIST), Tsukuba, Ibaraki 305-8565, Japan ‡

S Supporting Information *

ABSTRACT: We carried out a comprehensive study on the generality, scope, limitations, and mechanism of the palladium-catalyzed hydrophosphorylation of alkynes with P(O)−H compounds (i.e., H-phosphonates, H-phosphinates, secondary phosphine oxides, and hypophosphinic acid). For H-phosphonates, Pd/dppp was the best catalyst. Both aromatic and aliphatic alkynes, with a variety of functional groups, were applicable to produce the Markovnikov adducts in high yields with high regioselectivity. Aromatic alkynes showed higher reactivity than aliphatic alkynes. Terminal alkynes reacted faster than internal alkynes. Sterically crowded H-phosphonates disfavored the addition. For H-phosphinates and secondary phosphine oxides, Pd/dppe/Ph2P(O)OH was the catalyst of choice, which led to highly regioselective formation of the Markovnikov adducts. By using Pd(PPh3)4 as the catalyst, hypophosphinic acid added to terminal alkynes to give the corresponding Markovnikov adducts. Phosphinic acids, phosphonic acid, and its monoester were not applicable to this palladium-catalyzed hydrophosphorylation. Mechanistic studies showed that, with a terminal alkyne, (RO)2P(O)H reacted, like a Brønsted acid, to selectively generate the α-alkenylpalladium intermediate via hydropalladation. On the other hand, Ph(RO)P(O)H and Ph2P(O) H gave a mixture of α- and β-alkenylpalladium complexes. In the presence of Ph2P(O)OH, hydropalladation with this acid took place first to selectively generate the α-alkenylpalladium intermediate. A subsequent ligand exchange with a P(O)H compound gave the phosphorylpalladium intermediate which produced the Markovnikov adduct via reductive elimination. Related intermediates in the catalytic cycle were isolated and characterized. limited.7 Traditionally, the reaction of a halide RX with a phosphite (RO)3P at high temperatures (Michaelis−Arbuzov reaction) (Scheme 1, eq 1)8 and the addition of a P(O)−H

1. INTRODUCTION Because of the unique biological activity, organophosphorus compounds containing a phosphoryl P(O) group (phosphonates, phosphinates, and phosphine oxides) have great applications in medicinal and agricultural chemistry.1 For example, fosfomycin is a clinically used antibiotic,2a and glyphosate ((HO)2P(O)CH 2 NHCH 2 CO 2 H) 2b and glufosinate (MeP(O)(OH)CH2CH2CH(NH2)CO2H)2c are widely employed as herbicides. These organophosphoryl compounds are also widely used in organic synthesis.1a,h,i,3 For instance, phosphonates are extensively used in Horner−Wadsworth−Emmons reactions for the selective preparation of olefins,4a,b while phosphine oxides are good precursors to trivalent phosphines (including the chiral ones such as BINAP), which play a pivotal role as ligands in metalcatalyzed reactions.4c In addition, they also have wide applications in material chemistry.5 Because a P(O) group is able to ligate to a variety of metal ions and acids, phosphoryl compound-based metal extractants (especially those for U and Pu) are well known.5c,6a,b Fire retardancy is another unique feature of organophosphoryl compounds, and the development of an environment-benign phosphoryl compound-based fire retardant is of current concern.5d,e,6c,d Despite the importance of organophosphoryl compounds, general and efficient methods for their preparation were rather © 2018 American Chemical Society

Scheme 1. Commonly Employed C−P(O) Bond-Forming Reactions

bond to an olefin9 were the well-employed C−P(O) bondforming reactions. However, these reactions were usually only effective for the generation of a P−Csp3 bond. To solve this problem, the formation of a P−Csp2 bond via metal-catalyzed Received: January 15, 2018 Published: February 8, 2018 3139

DOI: 10.1021/jacs.8b00550 J. Am. Chem. Soc. 2018, 140, 3139−3155

Article

Journal of the American Chemical Society coupling of an alkenyl halide or ArX with a P(O)−H bond was then developed.7a,10 However, the preparation of an alkenylphosphoryl compound by this method was often not practicable due to the difficult accessibility to alkenyl halides. Metal-catalyzed addition of an H−heteroatom bond to a carbon− carbon unsaturated bond is one of the most straightforward and atom-efficient ways for the construction of a carbon−heteroatom bond (Scheme 2).11 As represented by hydrosilylation and

Scheme 3. Different Reactivity and Selectivity with P(O)−H Compounds

Scheme 2. Metal-Catalyzed Selective P(O)−H Addition to Alkynes (Hydrophosphorylation)

(Pd−P(O)), and what are the reactive intermediates involved in the catalytic cycles? In addition, since the reaction was only briefly communicated,16 information regarding generality, scope, and limitations, which is pivotal to synthetic chemists, is not available. To celebrate its 20th anniversary of discovery, as the discoverer of the reaction, herein, we report a full study on the palladiumcatalyzed hydrophosphorylation of alkynes with P(O)−H compounds. The results not only demonstrated the synthetic generality of the reaction (Scheme 4) but also answered the mechanistic puzzles associated with the reaction.20,21

hydroboration reactions catalyzed by metals, today these addition reactions have tremendous applications in both the laboratory and industry (Scheme 2). Hydrogen phosphoryl compounds (H-phosphonate (RO)2P(O)H, H-phosphinate (RO)RP(O)H, and H-phosphine oxide R2P(O)H) are a class of unique compounds. First, a tautomerism exists between the P(V) and P(III) forms [P(O)HP(OH)]. Therefore, they can ligate, like phosphines R3P, to transition metals.12 Second (and synthetically importantly), being different from the dangerous H-phosphines R2PH and phosphine halides R2PX, P(O)H compounds are rather air- and moisture-stable. They do not stink either. Third, these chemicals are readily available; i.e., they can be easily prepared or purchased since some (RO)2P(O)H are industrially manufactured. Therefore, the metalcatalyzed selective addition of P(O)−H compounds (hydrophosphorylation)13 is no doubt an ideal way for the preparation of alkenylphosphoryl compounds which are extremely useful14 but difficult to prepare by conventional methods.15 However, different from their silicon and borane counterparts, P(O)−H compounds can coordinate to metals in their P(III) forms P(OH) and significantly deactivate the catalyst.12 As a result, in a real sense, an efficient metal-catalyzed addition of (RO)2P(O)H to carbon−carbon unsaturated bonds under mild conditions was only realized in 1996 by employing a rather uncommon Me2Pd(PPh2Me)2 complex as the catalyst.16 Since then, this field has grown rapidly, and a variety of alkenylphosphoryl compounds can now be prepared by metal-mediated regio- and stereoselective hydrophosphorylation reactions.7f,17 Because of their novel applications, some compounds (vinylphosphonates CH2CHP(O)(OR)2, for example) are manufactured on industrial scale.18 However, despite more than 20 years passed since the first palladium-catalyzed hydrophosphorylation,16a a full study on this reaction has not been conducted yet, and a lot of puzzles remain unelucidated.7f For example: (1) Why does the palladiumcatalyzed addition of hydrogen phosphonates (RO)2P(O)H produce predominantly the Markovnikov adducts while (RO)PhP(O)H and Ph2P(O)H give a mixture of Markovnikov and anti-Markovnikov adducts?7f (2) Why can a trace amount of Ph2P(O)OH produce the Markovnikov adducts with higher selectivity (Scheme 3)?7f,19 (3) How does the palladium complex work, i.e., hydropalladation (Pd−H addition) vs phosphorylpalladation

Scheme 4. Palladium-Catalyzed Selective Hydrophosphorylation of Alkynes with P(O)−H Compounds

2. RESULTS AND DISCUSSION 2.1. Palladium-Catalyzed Hydrophosphorylation of Alkynes with H-Phosphonates (RO)2P(O)H.22 2.1.1. The Efficiency of the Palladium Catalyst. The early communication on the addition of (RO)2P(O)H to alkynes in the presence of a palladium catalyst provided a new way for the preparation of alkenylphosphonates under mild reaction conditions.16a However, the reaction conditions used are less practical since a rather special complex Me2Pd(PPh2Me)2 and too much of the catalyst were used. Therefore, we decided to optimize the reaction conditions.20 2.1.1.1. Effect of the Phosphine Ligand. We first conducted a catalyst’s screening by employing electronically and sterically different phosphines (Table 1). To minimize the effect of other factors, all these reactions were carried out by heating up an equimolar mixture of 1-octyne 1-1 with (MeO)2P(O)H 2-1 in the presence of 0.5 mol% PdMe2(PR3)2.23 The efficiency of the catalysts strongly depends on the phosphine ligands used. As shown in Table 1, a tiny change of the structure of the phosphine could lead to a dramatic change (up or down) of the yields of the products. Among the phosphines 3140

DOI: 10.1021/jacs.8b00550 J. Am. Chem. Soc. 2018, 140, 3139−3155

Article

Journal of the American Chemical Society

the adduct in high yields and selectivity. The reaction also proceeded in a toluene solution (runs 14 and 15) to give good yields of the adducts. 2.1.1.3. Effect of Impurities (H2O and Ph2P(O)OH). During the course of the study, we occasionally found that the reaction could become considerably slow with carefully dried, freshly distilled materials. Since commercially supplied (MeO)2P(O)H usually contains water, we thought it might be water that made such a difference. This is right. As shown in Table 2,

Table 1. Hydrophosphorylation of 1-Octyne with (MeO)2P(O)H Catalyzed by PdMe2(PR3)2a

Table 2. Effect of Water on the Hydrophosphorylation of 1-Octyne with (MeO)2P(O)H Catalyzed by PdMe2(PR3)2a

a

All reactions were similarly conducted in a sealed tube using an equimolar (MeO)2P(O)H and 1-octyne (0.5 mmol) with or without a solvent (toluene) in the presence of 0.5 mol% the palladium catalyst. b Determined by 1H NMR. α/β = the ratio of the α-adduct (the branched one) to the β-adduct (the linear one) (this notation also applies to other tables and schemes in this article). c0.25 mol% Pd catalyst. d0.1 mol% Pd catalyst. e2.5 M toluene solution. f0.8 M toluene solution. g0.1 mol% Pd catalyst. h0.05 mol% Pd catalyst.

a

All reactions were similarly conducted in a sealed tube using an equimolar (MeO)2P(O)H and 1-octyne (0.5 mmol) dissolved in 0.5 mL of toluene in the presence of 3 mol% palladium catalyst. b Determined by NMR. c3 mol% Ph2P(O)OH was used.

hydrophosphorylation of 1-octyne with (MeO)2P(O)H using either PdMe2(PPh2Me)2 or PdMe2(dppp) as a catalyst was affected by water very much. With PdMe2(PPh2Me)2, in the absence of water,24 the reaction gave 46% and 71% yields of the adducts, respectively, after 5 and 10 h (runs 1 and 2). However, the addition of a small amount of water accelerated the reaction. Just 3 mol% H2O (run 3) gave 88% yield of the adducts after 6 h, and 12 mol% H2O (run 5) gave a quantitative yield of the adducts after only 1.5 h! However, it was not that the more the water the higher the yield, because with 36 mol% water (runs 7 and 8), only 45% yield of the adducts was obtained after 10 h. Therefore, a small amount of water accelerates the reaction, but too much water retards the reaction. Water also improves the regioselectivity. This phenomenon was also observed with PdMe2(dppp) (runs 9−14). Ph2P(O)OH could also accelerate the reaction (run 15). 2.1.1.4. A Large-Scale Reaction. By using the optimized reaction conditions, the hydrophosphorylation reaction could be easily carried out in a relatively large scale. As demonstrated in Scheme 5, the two starting materials (200 mmol of 1-octyne and 200 mmol of (MeO)2P(O)H) were simply mixed together without purification25 and heated at 100 °C in the presence of a tiny amount of palladium catalyst (∼0.05−0.5 mol%). The corresponding adducts were obtained in high yields with high regioselectivity (97−98% selectivity) (Supporting Information). In addition to PdMe2dppp, the commercially available catalyst precursors Pd2(dba)3/dppp and Pd(OAc)2/dppp also worked efficiently.

investigated, PPh2Me is the best monodentate ligand while dppp and dppb are the best bidentate ligands. Thus, while PPh3 (run 1) only gave 10% yield of the products, PPh2Me could dramatically improve the catalyst’s efficiency to produce 78% yield of the adducts (run 2). However, surprisingly, a bulky alkyl group such as i-Bu (run 3) or cyclohexyl (run 4), produced a trace amount of the adducts. With PPhMe2 and PEt3, the addition proceeded only sluggishly (runs 5 and 6). A similar phenomenon was observed with bidentate phosphine ligands (runs 7−24). Thus, dppe (run 7, Ph2P(CH2)2PPh2) did not produce the adducts, while Ph2P(CH2)nPPh2 (run 10, dppp, n = 3; run 20, dppb, n = 4) could give excellent yields of the products with high regioselectivity. However, the yields of the adducts decreased as the methylene chain of the bisphosphine ligand became longer (runs 21 and 22). Other bidentate phosphines such as dppf (run 23) and BINAP (run 24) could also produce the adducts. With dppp, the reaction could proceed efficiently with 0.25 mol% catalyst (run 11), and even 0.1 mol% of the catalyst could also produce the adducts in 77% yield after 20 h (run 12), and an almost quantitative yield after 48 h (run 13). 2.1.1.2. Effect of Temperature and Concentration. Although the addition proceeded efficiently at 80 °C (Table 1, run 10), it progressed slowly at 60 and 70 °C (runs 8 and 9). When the reaction was conducted at elevated temperatures of 100 °C (run 16) and 120 °C (run 19), the reaction took place rapidly to produce 3141

DOI: 10.1021/jacs.8b00550 J. Am. Chem. Soc. 2018, 140, 3139−3155

Article

Journal of the American Chemical Society

Scheme 7. A Competitive Reaction of para-Substituted Phenylacetylenes

Scheme 5. 200-mmol-Scale Reactions for Aromatic and Aliphatic Alkynes

Table 3. A Competitive Reaction of H-Phosphonates: Yield of 3 vs Timea

Therefore, ca. 40 g of 3-1 could be readily prepared (eq 4 in Scheme 5). Similarly, a mixture of phenylacetylene and (MeO)2P(O)H gave the corresponding alkenylphosphonate 3-2 in 91% isolated yield (eq 5 in Scheme 5) (Caution!).25 2.1.2. Reactivity of Alkynes. 2.1.2.1. Aliphatic Alkynes vs Aromatic Alkynes. To compare the reactivity between an aliphatic alkyne and an aromatic alkyne, a competitive reaction between 1-octyne and phenylacetylene was carried out (Scheme 6).

a

Reaction conditions: 0.5 mmol of 1-octyne, an equimolar mixture of (RO)2P(O)H (0.1 mmol for each P(O)−H compound), 0.5 mol% Pd2(dba)3, 1 mol% dppp, 0.5 mL of toluene-d8, 80 °C. NMR yields based on (RO)2P(O)H.

Scheme 6. Reactivity of 1-Octyne vs Phenylacetylene

H-phosphonates followed a decreasing order of 2-5 > (MeO)2P(O)H > (EtO)2P(O)H = (n-BuO)2P(O)H > (i-PrO)2P(O)H. The difference of (RO)2P(O)H in reactivity can be simply attributed to the difference of their bulkiness, i.e., the smallest 2-5,27 reacted the fastest, and the most crowded (i-PrO)2P(O)H reacted the slowest. In fact, with a very bulky (AdO)2P(O)H (Ad = 1-adamantyl), no addition took place at all under similar conditions. The additions to phenylacetylene and 1-octyne (representatives for terminal aromatic alkynes and terminal aliphatic alkynes, respectively) are summarized in Table 4. As to (RO)2P(O)H

A mixture of (MeO)2P(O)H (0.5 mmol), 1-octyne (0.25 mmol), and phenylacetylene (0.25 mmol) in toluene (0.5 mL) was heated at 100 °C for 3 h in the presence of a palladium catalyst. The α-adducts 3-1 and 3-2 were generated in a ratio of 1:1.9. This result indicates that phenylacetylene reacts faster (ca. twice) than 1-octyne. 2.1.2.2. Steric Effect. As might be expected, a sterically bulky alkyne might favor the formation of the β-adduct.26 This is true. Although the above-mentioned 1-octyne and phenylacetylene selectively gave the α-adducts, under similar conditions, the steric bulky 3,3-dimethyl-1-butyne gave the adducts in 92% yield with a ratio α/β = 86/14 (eq 6).

Table 4. Palladium-Catalyzed Addition of H-phosphonates to Phenylacetylene and 1-Octynea

2.1.2.3. Electronic Effect. The electronic effect on the reactivity was studied by conducting a competitive reaction of phenylacetylenes with (MeO)2P(O)H (acetylene 0.1 mmol each, (MeO)2P(O)H 0.4 mmol, toluene 0.5 mL) (Scheme 7). All gave the α-adduct selectively, and the reactivity roughly follows an increasing order of H = Me < OMe < CF3. 2.1.3. Reactivity of H-phosphonates. Although no big difference in regioselectivity (all gave the α-adduct predominantly), as shown in Table 3, a strong steric effect was observed for (RO)2P(O)H; i.e., the bulkier the R, the lower the reactivity of (RO)2P(O)H. Competitive reactions of (RO)2P(O)H (R = Me, Et, n-Bu, i-Pr) and 2-5 with 1-octyne showed that the reactivity of

a

Reaction conditions: 0.5 mmol of H-phosphonate, 0.5 mmol of alkyne, 0.5 mol% Pd2(dba)3, 1 mol% dppp, 0.5 mL of toluene, 100 °C, overnight. bPd(PPh3)4 was used as the catalyst. c1 mmol of phenylacetylene was added, and 50% yield of 5 was produced. 3142

DOI: 10.1021/jacs.8b00550 J. Am. Chem. Soc. 2018, 140, 3139−3155

Article

Journal of the American Chemical Society (where R = Me, Et, n-Bu, Bn, and Ph) and the five-membered cyclic 2-5, all could be used as the substrates to produce the corresponding α-adducts selectively in high yields. As mentioned above, the bulky (i-PrO)2P(O)H is an exception (vide inf ra), which only gave 30% yield of 3-8 with phenylacetylene (run 7). However, interestingly, with 1-octyne, it could produce the corresponding adduct 3-12 in a high yield (run 12). Careful analysis of the products with phenylacetylene (run 7) revealed that, in addition to the mono-addition product 3-8, a butadiene 5 was generated in 50% yield (run 7). Although the exact pathway for the formation of 5 was not clear, a stepwise path, first the generation of an enyne 4 by a head-to-tail dimerization of phenylacetylene,28 followed by the addition of (i-PrO)2P(O)H to this enyne, might account for its formation (eq 7). In support of this hypothesis, a separate reaction using enyne 4 did produce 5 regioselectively (eq 7).

phenylacetylene and 1-octyne to produce the corresponding adducts in 85% and 87% yields, respectively (Table 4, runs 4 and 11). 2.1.4. Reactions with Diynes. Diynes also react similarly to give the adducts. However, interestingly, they can generate different kinds of products with different length of the (CH2)n linkage (Table 5). For example, hepta-1,6-diyne reacted with Table 5. Palladium-Catalyzed Hydrophosphorylation of Diynes with (MeO)2P(O)Ha

Therefore, there are two competitive reactions: the dimerization of phenylacetylene and the addition of P(O)H to phenylacetylene (Scheme 8). With the bulky (i-PrO)2P(O)H, the addition of a Reaction conditions: 0.5 mmol of (MeO)2P(O)H, 0.25 mmol of alkynes, 0.5 mol% Pd2(dba)3, 1 mol% dppp, 0.5 mL of toluene, 100 °C, overnight. bIsolated yield. c15% yield of the cyclized product 6 was produced. d0.25 mmol of (MeO)2P(O)H was used.

Scheme 8. Hydrophosphorylation vs Dimerization

(MeO)2P(O)H smoothly to afford, not the expected addition product, but a six-membered cyclic 3-15 in 95% yield (run 1). With octa-1,7-diyne, the normal addition product 3-16 was produced in 80% yields, and a seven-membered cyclic product 6 was generated in 15% yield (run 2). With nona-1,8-diyne, however, the normal hydrophosphorylation proceeded smoothly to give the addition product 3-17 in 91% yield, and the expected eightmembered cyclic product was not detected (run 3). Deca-1,9diyne reacted similarly to produce the corresponding addition product 3-18 in 93% yield (run 4). With deca-1,5-diyne, a monoaddition product 3-19 was obtained in 83% yield, and no cyclic phosphonates were detected (run 5).29 2.1.5. Scope and Limitations of the Palladium-Catalyzed Hydrophosphorylation of Alkynes with H-Phosphonates. To further disclose the scope and limitations of this palladiumcatalyzed hydrophosphorylation reaction, the reactions of a wide range of alkynes and hydrogen phosphonates (RO)2P(O)H were conducted (Table 6). A variety of alkynes, i.e., aromatic and aliphatic, terminal and internal, electron-rich and electron-poor, all were readily hydrophosphorylated to produce the corresponding alkenylphosphonates in high yields with high chemo- and stereoselectivity. The reaction features wide tolerance to a broad range of functional groups (i.e., the labile carbonyl, cyano, hydroxyl, halogens, free amine, anhydride, and even NO2 groups). Alkenylphosphonates of ferrocene and thiophene derivatives could also be prepared in high yields. As judged from the trans-adducts with internal alkynes, this metal-catalyzed addition of a P(O)−H bond to the triple bond took place via syn addition to give the trans-adducts. As to the

(i-PrO)2P(O)H to phenylacetylene was slow, and a considerable amount of phenylacetylene dimerized to generate 4 which eventually produced 5. The addition of (i-PrO)2P(O)H with 1-octyne could produce the corresponding alkenylphosphonate in a good yield because the dimerization of 1-octyne is slow compared to that of phenylacetylene.28 In addition to the above-mentioned steric effect, the electronic factor also significantly affects the addition. (CF3CH2O)2P(O)H 2-6, is such an example (for discussions, see sections 2.6.2 and 2.6.3). As shown in Scheme 9, (CF3CH2O)2P(O)H did not Scheme 9. Palladium-Catalyzed Hydrophosphorylation of 1-Octyne with (CF3CH2O)2P(O)H

efficiently produce the adducts using the Pd/dppp catalyst. Another catalyst Pd/PPh2Me that is effective for the hydrophosphorylation (Table 1) did not work either. Interestingly, by using Pd(PPh3)4, (CF3CH2O)2P(O)H reacted with both 3143

DOI: 10.1021/jacs.8b00550 J. Am. Chem. Soc. 2018, 140, 3139−3155

Article

Journal of the American Chemical Society Table 6. Palladium-Catalyzed Hydrophosphorylation of Alkynes with Hydrogen Phosphoryl Compoundsa

a Reaction conditions: For H-phosphonates: 0.5 mmol of H-phosphonate, 0.5 mmol of alkynes, 0.5 mol% Pd2(dba)3, 1 mol% dppp, 0.5 mL of toluene, 100 °C, 4 h to overnight. For H-phosphinates and H-phosphine oxides: 0.5 mmol of H-phosphonate, 0.5 mmol of alkynes, 1 mol% Pd2(dba)3, 2 mol% dppe, 4 mol% Ph2P(O)OH, 0.5 mL of toluene, 100 °C, overnight. For hypophosphinic acid: 10 mmol of alkynes, 10 mmol of H3PO2, 2 mol% Pd(PPh3)4, 10 mL of THF, room temperature, overnight. b1 mol% Pd2(dba)3, 2 mol% dppp, 0.5 mL of dioxane. c2 mol% Pd(OAc)2, 3 mol% dppp, 0.5 mL of dioxane. d3 equiv (MeO)2P(O)H (1.5 mmol) was used. With 1 equiv (MeO)2P(O)H, the yield is ca. 51% due to polymerization of the alkyne. eRatio of the two regioisomers. f1 atm acetylene gas. gNMR spectroscopically pure mixture of an H-alkenylphosphinate with its reduced form in a ratio shown in the parentheses.

regioselectivity, the addition well follows the Markovnikov’s rule to give the corresponding adducts selectively (Scheme 10). Therefore, a terminal alkyne with an electron-donating group (EDG)

produces the branched adduct while that with an electronwithdrawing group (EWG) produces the linear one. For an internal alkyne having one EDG and one EWG, the phosphoryl 3144

DOI: 10.1021/jacs.8b00550 J. Am. Chem. Soc. 2018, 140, 3139−3155

Article

Journal of the American Chemical Society

complexes with a more basic phosphine PEt3 and PMe3 did not catalyze the addition satisfactorily. Remarkably, when a combination of Me2Pd(PPhMe)2 and Ph2P(O)OH was employed, an excellent yield of the α-adduct 3-62 was obtained with a high regioselectivity. A combination of 1 mol% Pd2(dba)3, 2 mol% dppe, and 4 mol% Ph2P(O)OH, could also give a good result (run 8). As shown in Table 6, the reaction with Ph(EtO)P(O)H has a wide generality, and the Markovnikov adduct was generated selectively. Thus, with the exception of trimethylsilylacetylene which gave the β-trans-adduct 3-68 selectively, both aliphatic and aromatic terminal alkynes tested all reacted efficiently, affording the α-adducts selectively. In addition, a variety of functionalities such as chloro (3-65), cyano (3-64), ester (3-67), silyl (3-68), alkenyl (3-66), and thienyl (3-71) groups were well tolerant. Two P(O) groups were also easily introduced into nona-1,8-diyne (3-69). Though prolonged heating was needed compared with terminal alkynes, an internal alkyne like tolane could also be successfully hydrophosphorylated, producing the corresponding trans-adduct selectively (3-73). Acetylene gas was also hydrophosphorylated to give the corresponding vinylphosphinate 3-63 in 76% yield. 2.3. Palladium-Catalyzed Hydrophosphorylation of Alkynes with Secondary Phosphine Oxide R′RP(O)H.32 Addition of Ph2P(O)H 2-8 to terminal alkynes catalyzed by Me2Pd(PPhMe)2/Ph2P(O)OH selectively afforded the Markovnikov adducts.19a,33 In order to generalize this reaction, a screening on the catalyst was carried out (Table 8). The addition proceeded

Scheme 10. Regioselectivity of the Pd-Catalyzed Addition of (RO)2P(O)H to Alkynes Follows the Markovnikov’s Rule

group selectively bonds to the same carbon with the EDG. For trimethylsilylacetylene, the phosphoryl group selectively bonds to the terminal carbon selectively.30 However, limitations do exist with alkynes (Figure 1). For example, 1-(ethynylsulfonyl)-4-methylbenzene did not produce

Figure 1. Alkynes not applicable to the reaction.

the corresponding adduct because of the decomposition of the substrate under the present conditions. A few propargyl alkynes as shown in Figure 1 could not be hydrophosphorylated either. 2.2. Palladium-Catalyzed Hydrophosphorylation of Alkynes with H-Phosphinates R′(RO)P(O)H.31 H-Phosphinates (RO)R′P(O)H can also add to alkynes catalyzed by a palladium catalyst. However, interestingly, the best catalyst for their addition is dif ferent f rom that of H-phosphonates, despite the structures’ similarity. In order to elucidate the regio- and stereoselectivity of the addition of hydrogen phosphinates to alkynes, as a model reaction, we first investigated the addition of Ph(EtO)P(O) H 2-7 to 1-octyne (Table 7). None of the palladium catalysts tested

Table 8. Addition of Ph2P(O)H to Phenylacetylene Catalyzed by Palladiuma

Table 7. Pd-Mediated Addition of Ph(EtO)P(O)H to 1Octynea

a

Condition: 0.5 mmol of phenylacetylene, 0.5 mmol of diphenylphosphine oxide, 1 mol% Pd2(dab)3, ligand (Pd/P = 1:2), 4 mol% Ph2P(O)OH, and 0.5 mL of toluene were mixed, 100 °C, overnight.

smoothly to give the corresponding Markovnikov adduct using dppe as the ligand (run 1). The results compiled in Table 6 showed that, in general, H-phosphine oxides could add to the alkynes chemo- and regioselectively. With 1,4-bis(phenylphosphoryl)benzene, the corresponding adduct 3-83 was obtained in 87% yield. 1-Octyne also gave the alkenylphosphoryl oxides in high yields selectively (3-81, 3-82). However, a bulky tert-butylphenylphosphine oxide only gave a trace amount of the adduct (3-78). 2.4. Palladium-Catalyzed Hydrophosphorylation of Alkynes with Hypophosphinic Acid H2P(O)(OH).34 The addition of hypophosphinic acid 2-9 to alkynes could also take place in the presence of a proper palladium catalyst. As shown in Table 9, Pd(OAc)2 was reduced to the metallic palladium black, and no addition products were detected (run 1). No addition

a

Conditions: 0.5 mmol of 1-octyne, 0.5 mmol of Ph(EtO)P(O)H, 5 mol% catalyst, and 0.5 mL of toluene, 70 °C, 5 h. bThe yield and ratio were based on 31P NMR spectra. c1 mol% Pd2(dba)3 (Pd/P = 1/2), 100 °C, overnight.

was able to afford the corresponding adducts in satisfactory yields and selectivity. Thus, the common Pd(PPh3)4 only gave a trace amount of the product. Me2Pd(PPh3)2, Me2Pd(PPh2Me)2, and Me2Pd(PPhMe2)2 could produce the adduct 3-62 in moderate yields. However, the selectivity was not satisfactory. Palladium 3145

DOI: 10.1021/jacs.8b00550 J. Am. Chem. Soc. 2018, 140, 3139−3155

Article

Journal of the American Chemical Society

be hydrophosphorylated by hypophosphinic acid, producing the corresponding H-alkenylphosphinic acids in good yields with high regioselectivity. As described in Supporting Information, a very pure Halkenylphosphinic acid is difficult to isolate by conventional SiO2-column chromatography techniques because of the presence of the P(O)(OH) unit. However, as shown in Scheme 11,

Table 9. Addition of Hypophosphinic Acid to 1-Octyne Catalyzed by Palladium

Scheme 11. Hydrophosphorylation of Phenylacetylene with H2P(O)(OH) at 10 mmol Scale and Its Purification

because of the high efficiency of the reaction, NMR spectroscopically pure H-alkenylphosphinic acid could be generated easily. 2.5. Hydrogen Phosphoryl Compounds Not Applicable to the Reaction. As described above, in addition to the esters of H-phosphonates, H-phosphinates, and secondary phosphine oxides, hypophosphinic acid could be used as the hydrogen phosphoryl compounds in the palladium-catalyzed addition of P(O)−H bonds to alkynes to produce the corresponding alkenylphosphorus compounds efficiently. However, a hydrogen phosphoryl compound bearing a free OH group (with the exception of H3PO2), i.e., phosphonic acid and phosphinic acids, could not be used in the reaction, and no addition products were produced under similar conditions. (Figure 2; see section 2.6.7 for discussions).

a Molar ratio: alkyne/P(O)H = 2.5/1. bDry H3PO2. cMolar ratio: alkyne/P(O)H = 1/1. d2 mol% Pd catalyst, 1 M solution.

Figure 2. Hydrogen phosphoryl compounds applicable and not applicable to the Pd-catalyzed hydrophosphorylation of alkynes.

took place either with Pd2(dba)3 or Pd(PCy3)2 (runs 2 and 3). Interestingly, however, the addition took place when Me2Pd(PPh3)2 was used as the catalyst, and 75% yield of 3-84 was generated after 24 h (runs 4−6). A small amount of 9 (3-84/9 > 95/5) via the reduction of 3-84 with H3PO2 was also detected.35 Surprisingly, a similar addition did not occur with other palladium complexes (runs 7−9). Phosphine ligands PPh2R (R = Me, cyclohexanyl) were also effective for this reaction (runs 10−13). As to the bidentate ligand dppp, which is effective for the addition of (MeO)2P(O)H as described above, only gave a low yield of the product (run 14). On the other hand, dppf gave a quantitative yield of the adduct (run 21). Interestingly, an excess amount of 1-octyne slowed down the reaction (run 22), and only a low yield of the adduct was obtained with a dried H3PO2 (run 23). The common Pd(PPh3)4 or a combination of Pd(OAc)2 (or Pd2(dba)3) with PPh3 also well catalyzed this addition (runs 24−31). With 2 mol% Pd(PPh3)4, 95% yield of the adduct could be obtained (run 33). The results compiled in Table 6 showed that the addition of hypophosphinic acid to alkynes was a general reaction. Under the conditions of run 33, both aromatic and aliphatic alkynes could

2.6. Mechanistic Insights. 2.6.1. Reactions Associated with Me2Pd(PR3)2 in the Catalytic Reactions. As reported earlier, dimethylpalladium complex cis-Me2Pd(PPh2Me)223 was initially used as a catalyst for the addition of (MeO)2P(O)H to alkynes.16a However, details about how this divalent Pd(II) complex catalyzed the addition were not known. To clarify the mechanism, several stoichiometric reactions were conducted. cis-Me2Pd(PPh2Me)2 10-1 (117.6 mg, 0.219 mmol) and (MeO)2P(O)H 2-1 (0.482 mmol) were dissolved in toluene (1.5 mL) at room temperature to give a transparent solution (Scheme 12). White solids gradually precipitated out from the solution which was determined to be the trans-[(MeO)2P(O)]2Pd(PPh2Me)2 12-1 (129 mg, 81% isolated yield).36 Although a pure monomethylpalladium complex 11-1 was not isolated, as described below, the formation of complex 12-1 was considered to take place through a stepwise protonolysis of cis-Me2Pd(PPh2Me)2 with (MeO)2P(O)H. 3146

DOI: 10.1021/jacs.8b00550 J. Am. Chem. Soc. 2018, 140, 3139−3155

Article

Journal of the American Chemical Society

Complexes 12-1 and 12-2 are rather thermally stable since no decomposition was observed after their benzene solution was heated at 80 °C for 5 h. Interestingly, however, they readily reacted with phenylacetylene (1 equiv) at room temperature via ligand exchange to liberate (MeO)2P(O)H (Scheme 13). This reaction seems reversible since the four chemicals were present in the mixture even after 24 h, and no further formation of 13 was observed when the ratio of 12/13 reached ca. 1/1. Upon heating the mixture (50−67 °C), however, all these chemicals disappeared, and a new complex 14-1 was formed quantitatively. With PEt3, however, no corresponding olefin-coordinated complex could be detected under the current conditions. No further reaction of complex 14-1 with an additional (MeO)2P(O)H (10 equiv) was observed at 67 °C. However, as phenylacetylene (10 equiv) was subsequently added, the addition of (MeO)2P(O)H to phenylacetylene took place to produce the adduct 3-2 in 88% yield after 3 h. As confirmed by NMR spectroscopy, complex 14-1 was also clearly observed in the mixture after the reaction. These results suggested that dimethylpalladium complex 10-1 only served as a catalyst precursor, which should change to zerovalent Pd(0) species through thermal decomposition of 10-123 and/or a series of reactions as described above. These zerovalent Pd(0) species acted as the catalyst for the hydrophosphorylation reaction. 2.6.2. Reaction of (RO)2P(O)H with Pd(PEt3)4 and Phenylacetylene: H−Pd Addition vs P(O)−Pd Addition and the Acetylene Path vs the H−Pd Path. Oxidative addition of (RO)2P(O)H to Pd(PEt3)4 could generate the corresponding Pd−H species (Scheme 14). However, the Pd−H complex is

Scheme 12. Reactions of Dimethylpalladium Complexes with P(O)−H Compounds

By using a more stable cis-Me2Pd(PEt3)2 10-2,23 the stepwise protonolysis process could be confirmed clearly. Thus, at room temperature, (MeO)2P(O)H (0.267 mmol) was added to the solution of cis-Me2Pd(PEt3)2 (0.089 mmol) in C6D6 (0.5 mL). Gas evolution was observed immediately. A new signal assigned to 11-2 emerged at δ 93.9 (P(O), t, JPP = 50.0 Hz, 1P) and δ 21.0 (Et3P, d, JPP = 50.0 Hz, 2P) in 31P NMR spectroscopy. As estimated from 31P NMR spectroscopy, the ratio of 11-2 vs cis-Me2Pd(PEt3)2 was 58/42 and 10/90 after 1.5 and 3.5 h, respectively. The monomethylpalladium complex 11-2 was rather stable under the current conditions, i.e., no thermal decomposition of 11-2 to MeP(O)(OMe)2 was observed, and further protonolysis of 11-2 with (MeO)2P(O)H to form 12-2 (P(O), 81.0 ppm) was also negligible. However, further protonolysis of 11-2 with (MeO)2P(O)H did take place at a slightly elevated temperature. Thus, heating the solution at 60 °C for 3.5 h resulted in a complete disappearance of 11-2 and a quantitative formation of 12-2 (95% isolated yield). A separate experiment confirmed that the addition of free PEt3 did not retard the protonolysis. However, interestingly, the two configuration isomers of Me2Pd(PEt3)2 showed different reactivity toward (MeO)2P(O)H, i.e., trans-Me2Pd(PEt3)2 was more reactive than the cis-Me2Pd(PEt3)2 complex. For example, monitoring the reaction of Me2Pd(PEt3)2 (cis/trans = 37/63, 33 mg, 0.089 mmol) with (MeO)2P(O)H (3 equiv) in C6D6 (0.5 mL) by 31P NMR spectroscopy showed that the trans-Me2Pd(PEt3)2 complex completely disappeared within 1 h, and ca. 15% of cisMe2Pd(PEt3)2 remained.37 More interestingly, the above reaction was considerably accelerated by a trace amount of water. Thus, two reactions of cisMe2Pd(PEt3)2 (0.0693 mmol) with (MeO)2P(O)H (2.2 equiv) in C6D6 (0.4 mL) were carried out at room temperature, one had additional H2O (1.0 μL) and the other had not. The reaction of cis-Me2Pd(PEt3)2 with water was completely consumed to give the corresponding palladium complexes, while 36% of cis-Me2Pd(PEt3)2 remained without additional water after 80 min. This was consistent with the result in the catalytic reaction that a trace amount of water accelerated the Me2Pd(PPh2Me)2-catalyzed addition of (MeO)2P(O)H to 1-octyne (see section 2.1.1.3, Effect of Impurities).38

Scheme 14. Reactions of P(O)−H Compounds with Pd(0) Complex

rather unstable and decomposes easily at room temperature. For example, addition of (MeO)2P(O)H (0.2 mmol) to Pd(PEt3)4 (0.1 mmol) in C6D6 (0.5 mL) at room temperature could produce Pd−H in ca. 60% yield with a trace amount of the further reaction product P(O)−Pd−P(O) 12-2 in 15 min. However, this

Scheme 13. Role of Bisphosphorylpalladium Complexes in the Hydrophosphorylation

3147

DOI: 10.1021/jacs.8b00550 J. Am. Chem. Soc. 2018, 140, 3139−3155

Article

Journal of the American Chemical Society

As shown in Table 10, the reaction of phenylacetylene (0.05 mmol), H-phosphonate (0.05 mmol), and Pd(PEt3)4 (0.05 mmol) in C6D6 (0.5 mL) at room temperature was followed by NMR. Both steric and electronic factors of the substituent RO significantly affected this reaction: a H-phosphonate with an EWG reacts faster, while that with a steric bulky substituent reacts slower. For example, with (CF3CH2O)2P(O)H, the reaction almost completed within 4 h (run 3). On the other hand, with (i-PrO)2P(O)H, the reaction required more than 4 days (run 4). These reactions with (RO)2P(O)H very much resembled that of carboxylic acids, that proceeds via the alkyne path (path a) rather than the hydride path (path b), to give the palladium complex with Pd attached to the internal carbon of the alkyne.40 We assume that (RO)2P(O)H might similarly react via the alkyne path (path a) to produce the palladium complex (Scheme 15).40 2.6.3. Reductive Elimination of Alkenyl(phosphoryl)palladium Complexes. Complexes 16 are the first successfully isolated, structurally unambiguously determined (alkenyl)(phosphoryl)palladium complexes from the hydropalladation of alkynes. As described below, their isolation provides important clues to the catalytic cycle. As shown in Table 11, although 16-2 is stable at room temperature, reductive elimination of the complex did take place at

Pd−H complex was not stable and further reacted with (MeO)2P(O)H to give the bisphosphorylpalladium complex 12-2 in 55% yield after 2 h. The more reactive five-membered P(O)−H compound produced the bisphosphorylpalladium complex predominantly, and only a trace amount of Pd−H species could be detected from the reaction. When (EtO)2P(O)H (0.05 mmol) was added to a mixture of Pd(PEt3)4 (0.05 mmol) and phenylacetylene (0.05 mmol) in C6D6, the color of the solution gradually turned from yellow to colorless. After 1 h, 1H NMR showed that, while most of (EtO)2P(O)H remained unreacted, the free phenylacetylene almost disappeared because of its coordination to Pd(PEt3)4. Only a trace amount of Pd−H was detected (eq 9). Gradually, two

characteristic vinyl proton signals at 6.50 ppm (d, JP−H = 30.4 Hz), 5.17 ppm (d, JP−H = 15.2 Hz), and two new phosphorus signals at 88.67 ppm (t, P(O), JP−P = 51.5 Hz) and 13.99 ppm (d, PEt3, JP−P = 51.5 Hz) appeared, which, as described later, were due to a vinyl(phosphoryl)palladium complex 16-2.39 The starting materials disappeared after 30 h. Removal of the volatiles in vacuo gave 16-2 as a colorless oil (Table 10, run 2).

Table 11. Reductive Elimination of Alkenyl(phosphoryl)palladium Complexesa

Table 10. Generation of Alkenylpalladiums from Phenylacetylene, Pd(PEt3)4, and H-Phosphonates

a

Reaction conditions: 0.05 mmol of complex 16 in 0.5 mL of dry and degassed C6D6 was heated at 100 °C until the starting material was consumed. The reaction process was monitored by 1H/31P NMR spectroscopies.

The above experiment in eq 9 revealed two facts regarding the reactivity and the selectivity of the palladium complexes: (1) formally, hydropalladation (H−Pd addition to the alkyne), rather than phosphorylpalladation (P(O)−Pd addition to the alkyne) took place; (2) hydropalladation took place predominantly with Pd moiety bond to the internal carbon of phenylacetylene.

an elevated temperature (100 °C, 6 h), to generate the corresponding alkenylphosphonate that was isolated as an olefin− Pd(0) complex 17-2. Compared with the hydropalladation reactions of alkynes (see section 2.6.2), where a significant steric effect was observed, the

Scheme 15. Hydropalladation Paths of Alkynes with H-Phosphonates

3148

DOI: 10.1021/jacs.8b00550 J. Am. Chem. Soc. 2018, 140, 3139−3155

Article

Journal of the American Chemical Society steric influence seems small with the reductive elimination since MeO and i-PrO all gave similar results (runs 1 and 4). However, a big electronic effect was observed.21a Thus, with CF3CH2O, the reductive elimination was slow compared to other RO groups (run 3). These results showed that the reductive elimination was slower and more difficult than the hydropalladation. Therefore, it is assumed that the reductive elimination step perhaps should be the rate-determining step in the palladium-catalyzed hydrophosphorylations.41 2.6.4. Reactions with H-Phosphinate Ph(EtO)P(O)H and H-Phosphine Oxide Ph2P(O)H. Very interestingly, the reaction is highly structure-sensitive with a P(O)−H compound, and the three compounds, i.e., H-phosphonates (RO)2P(O)H, H-phosphinates Ph(RO)P(O)H, and H-phosphine oxide Ph2P(O)H, showed quite different behaviors under similar conditions. H-phosphinate Ph(EtO)P(O)H 2-7 gradually reacted at room temperature. However, being different from (RO)2P(O)H, a mixture of alkenyl(phosphoryl)palladium complexes 16-5 and 16-5′ was generated. In addition, these complexes were not stable at room temperature; they decomposed slowly via reductive elimination to give a mixture of the corresponding olefin-palladium(0) complexes 17-5 and 17-5′ (eq 10). With a cyclic H-phosphinate 2-10, the alkenyl(phosphoryl)palladiums 16-6 and 16-6′ (generated in ca. 3:2 ratio) are stable (eq 11) and were successfully isolated and characterized (see Table 13).

Consequently, the observed different behavior of the three kinds of P(O)−H compounds ((RO)2P(O)H, Ph(EtO)P(O)H, and Ph2P(O)H) reflects the results of the two factors. The formation of the Markovnikov intermediate 16 is more electronically favored, while the formation of the anti-Markovnikov intermediate 16′ is more sterically favored. Considering that a carboxyl acid RCO2H can predominantly generate the Markovnikov intermediate,40 and the acidity of P(O)H compounds follows a decreasing order of (RO)2P(O)H > Ph(EtO)P(O)H > Ph2P(O)H,42 we assume that, the more acidic (RO)2P(O)H might react, like RCO2H, to predominantly give 16, while steric factors weighted more with the less acidic Ph(OEt)P(O)H and Ph2P(O)H which, consequently, gave a mixture of 16 and 16′ (Scheme 16). Scheme 16. “Acidity” Formally Determines the Regioselectivity of the Hydropalladation

2.6.6. Role of a Brønsted Acid. The addition of a trace amount of phosphinic acid Ph2P(O)OH could significantly enhance the reactivity and reverse the regioselectivity of the palladium-catalyzed addition of Ph2P(O)H with a terminal alkyne.19 However, the role of Ph2P(O)OH was not understood. 2.6.6.1. Regio- and Stereoselective Hydropalladation of an Alkyne with Ph2P(O)OH. Diphenylacetylene and phenylacetylene could be easily hydropalladated with the combination of carboxylic acid and Pd(PEt3)4.40 The reaction took place stereoselectively to produce the alkenylpalladium complexes in high yields (Scheme 17). With an aliphatic alkyne 1-octyne, the

With Ph2P(O)H 2-8, the corresponding alkenylpalladiums quickly decomposed to 17-6 (17-6′) at room temperature. Thus, only a trace amount of 16-7 (16-7′) could be detected by NMR from the reaction where the reductive elimination products 17-6 (17-6′) (17-6/17-6′ ≈ 1:4) were generated predominantly (eq 12). When Ph2P(O)H with a trace amount of Ph2P(O)OH

Scheme 17. Hydropalladation of Alkynes with the Combination of Pd(0) Complex and Carboxylic Acids

(ca. 5 mol%) was used, a mixture of 17-6 (17-6′) with a ratio of 17-6/17-6′ ≈ 11.5:1 was obtained (for the effect of Ph2P(O)OH, see section 2.6.6). Therefore, the reductive elimination of complexes 16 follows an increasing order of (RO)2P(O) < (RO)PhP(O) < Ph2P(O) (eq 13). 2.6.5. Explanation for the Difference in Reactivity between (RO)2P(O)H, Ph(EtO)P(O)H, and Ph2P(O)H. Both electronic and steric factors affect the reactivity of the P(O)−H compounds.

reaction also occurred to give the branched palladium complex regioselectively in a high yield. Brønsted acids (acetic acid and Ph2P(O)OH as described below) were more reactive than P(O)−H compounds in the hydropalladation of alkyne. With Ph2P(O)OH, a similar hydropalladation took place. Thus, as demonstrated in Table 12, diphenylacetylene was quickly hydropalladated generating the corresponding cis-alkenylpalladium 3149

DOI: 10.1021/jacs.8b00550 J. Am. Chem. Soc. 2018, 140, 3139−3155

Article

Journal of the American Chemical Society Table 13. Ligand Exchange Reactionsa

Table 12. Hydrometalation of Alkynes with a Combination of M(PEt3)4 and P(O)(OH) Acida

a Reaction conditions: 0.05 mmol of alkynes, 0.05 mmol of M(PEt3)4, 0.05 mmol of acid, and 0.5 mL of C6D6 were mixed in a glovebox at room temperature for 2 h. The complex was isolated by recrystallization from hexane−toluene at −30 °C (if solid). bThe complexes were colorless oil.

complexes 18-4 at room temperature in 96% yield (run 1). With phenylacetylene, a Markovnikov-adduct alkenylpalladium complex 18-5 was readily generated regioselectively (run 2). The structure of 18-5 was unambiguously confirmed by the X-ray analysis.28a Similarly, diphenylphosphoric acid (PhO)2P(O)OH also selectively gave the branched alkenylpalladium 18-6 in 85% yield (run 3). In addition, Ni(PEt3)4 was as reactive as Pd(PEt3)4 to afford the corresponding branched alkenylnickel complex 18-7 in 90% yield (run 4). As to the reactivity of an alkyne, an aromatic alkyne is more reactive than an aliphatic alkyne. For example, the reaction of an equimolar mixture of phenylacetylene (0.05 mmol) and 1-octyne (0.05 mmol) with Pd(PEt3)4 (0.05 mmol) and acetic acid (0.1 mmol) in toluene-d8 (0.5 mL) at room temperature, completed within 1 h to generate 18-2 and 18-3 in the ratio of 5.5:1 (eq 14).

a

Reaction conditions: 0.05 mmol of alkynes, 0.05 mmol of M(PEt3)4, 0.1 mmol of HOAc, and 0.5 mL of C6D6 were mixed at room temperature for 20 h. Then, 0.05 mmol of P(O)H compound was added. The mixture was kept at room temperature until the P(O)H compound was consumed. The complex was isolated by recrystallization from hexane−toluene at −30 °C (when solid). bThe complexes were oil. cToluene was used as solvent. dThose complexes contain HOAc. eRecrystallization from hexane−toluene at room temperature.

2.6.6.2. Rapid Ligand Exchange. At room temperature, these alkenylpalladium complexes 18 react with P(O)−H compounds (H-phosphonates, H-phosphinates, and H-phosphine oxides) quickly to give the corresponding alkenyl(phosphoryl)palladium complexes 16 in high yields via ligand-exchange reactions (Table 13). Thus, 0.05 mmol of alkenylpalladium complex 18-1 and 0.05 mmol of P(O)−H 2-5 were mixed in 0.5 mL of C6D6 at room temperature, as indicated by NMR spectroscopies, the starting materials 18-1 and 2-5 were consumed in 2 h, while a new characteristic alkenyl proton signal at 6.59 ppm (d, JP−H = 18.0 Hz) and two new phosphorus signals at 104.99 ppm (t, JP−P = 52.3 Hz), 9.29 ppm (d, JP−P = 52.3 Hz) were observed, indicative of the formation of a new alkenyl(phosphoryl)palladium complex, 16-9. Removal of the volatiles in vacuo afforded 16-9 as a white solid, which was recrystallized from toluene-hexane at −30 °C to give single crystals suitable for X-ray analysis (run 7). The ORTEP drawing of complex 16-9 was depicted in Figure 3, which unambiguously reveals the trans geometry of the carbon−carbon double bond as well as the trans geometry on Pd atom. The Pd atom in complex 16-9 adopts a square-planar coordination geometry, which is ligated by two PEt3 in a trans manner.

Figure 3. ORTEP drawing of complex 16-9. Thermal ellipsoids are drawn at 50% probability. H atoms and HOAc, which was attached to the PO by a hydrogen bond, are omitted for clarity. Selected bond lengths (Å) and angles (deg): C7−C8 = 1.335(2), C7−Pd1 = 2.0768(15), C7−C15 = 1.483(2), P2−Pd1 = 2.3492(5), P1−Pd1 = 2.3287(2); C8−C7−Pd1 = 126.41(12), C7−Pd1−P2 = 89.37(5), P1−Pd1−P2 = 88.882(16), P2−Pd1−P3 = 165.533(17).

Complex 18-2 could also react with H-phosphonate 2-5 efficiently to give the corresponding alkenyl(phosphoryl)palladium complexes 16-8 in 80% isolated yields at room temperature (run 5). This ligand exchange reaction was significantly retarded with sterically bulky P(O)−H compounds (runs 1−4), i.e., the bulkier the substrate, the slower the reaction. With diethyl phosphonate 2-2, the reaction required 9 h to give 92% yield of the complex. However, with diisopropyl phosphonate 2-3, the reaction took several days (for details, see Supporting Information). With H-phosphinate 2-10, the ligand exchange reaction also proceeded 3150

DOI: 10.1021/jacs.8b00550 J. Am. Chem. Soc. 2018, 140, 3139−3155

Article

Journal of the American Chemical Society

the reductive elimination product 17-6 at room temperature in 80% isolated yield (eq 16, Figure 6).

smoothly to afford the corresponding complex 16-6 in 85% isolated yield (run 6). The structures of complexes 16-6 and 16-8 were also unambiguously determined by X-ray analysis (Figures 4 and 5).

Figure 4. ORTEP drawing of complex 16-6. Thermal ellipsoids are drawn at 50% probability. H atoms are omitted for clarity. Selected bond lengths (Å) and angles (deg): C1−C2 = 1.336(2), C2−Pd = 2.0965(15), C2−C3 = 1.482(2), P2−Pd = 2.3240(5), P1−Pd = 2.3310(5); C1−C2−Pd = 118.72(12), C2−Pd−P2 = 90.23(4), P1−Pd−P2 = 88.779(15), P2−Pd−P3 = 174.047(15), O1−P1−Pd = 121.63(5).

Figure 6. ORTEP Drawing of complex 17-6. Thermal ellipsoids are drawn at 50% probability. H atoms are omitted for clarity. Selected bond lengths (Å) and angles (deg): C1−C2 = 1.431(8), C1−Pd1 = 2.090(6), C2−Pd1 = 2.162(6), P2−Pd1 = 2.2807(17), P1−Pd1 = 2.3291(19), C2−C3 = 1.508(9), C2−P3 = 1.781(6), O−P3 = 1.491(4); C1−Pd1−C2 = 39.3(2), C1−C2−Pd1 = 119.9(6), C1−Pd1−P2 = 99.69(16), C2−Pd1−P1 = 111.03(16), P1−Pd1−P2 = 109.71(6), C3−C2−Pd = 109.1(4), C2−P3− O = 116.1(3).

These elementary reactions can reasonably explain why in the presence of Ph2P(O)OH, the palladium-catalyzed addition of Ph2P(O)H to alkynes selectively produces the Markovnikov adducts,19 i.e., first, hydropalladation takes place selectively with Ph2P(O)OH to produce alkenylpalladium 18 which then reacts with Ph2P(O)H via a rapid ligand-exchange reaction to give the adduct 17 via an intermediate 16 (Scheme 18). Scheme 18. Role of Ph2P(O)OH Figure 5. ORTEP drawing of complex 16-8. Thermal ellipsoids are drawn at 50% probability. H atoms and a HOAc attached to the PO via a hydrogen bond are omitted for clarity. Selected bond lengths (Å) and angles (deg): C1−C2 = 1.330(3), C2−C3 = 1.482(3), C2−Pd = 2.081(2), P2−Pd = 2.3362(7), P1−Pd = 2.3337(7); C1−C2−Pd = 120.27(18), C2−Pd−P2 = 88.36(6), P1−Pd−P2 = 90.36(2), P2−Pd− P3 = 175.37(2).

As expected, complex 18-4 underwent similar ligand-exchange reaction with P(O)−H compounds to quantitatively give the phosphorylpalladium complex (eq 15).

2.6.7. P(O)−H Compounds Applicable to PalladiumCatalyzed Hydrophosphorylation of Alkynes: P(O)−H Activation vs O−H Activation. As described above, with the exception of hypophosphinic acid H3PO2, substrates with a hydroxyl OH group (P(O)OH) are not applicable to this palladium-catalyzed hydrophosphorylations (Figure 2). In order to understand this phenomenon, reactions of P(O)OH compounds with Pt(PEt3)4 were investigated (Scheme 19).43 As shown in Scheme 19, 0.05 mmol of H3PO2 (50% solution in water) reacted with 0.05 mmol of Pt(PEt3)4 in 0.5 mL of benzene-d6 quickly at room temperature to give a mixture of Pt−H complexes 15-3 (via the activation of an O−H bond) and 15-3′ (via the activation of P(O)−H bond) with a ratio of

When Ph2P(O)H was used as the substrate, the corresponding alkenylpalladium complex could not be isolated because of its rapid decomposition as described above (eq 12). Thus, 18-2 reacted with diphenylphosphine oxide 2-8 to produce not the corresponding alkenyl(phosphoryl)palladium complex 16-7, but 3151

DOI: 10.1021/jacs.8b00550 J. Am. Chem. Soc. 2018, 140, 3139−3155

Article

Journal of the American Chemical Society

The above results might indicate that except H3PO2, the reactive P(O)−Pd−H species could not be easily achieved with a P(O)(OH)H substrate, which may explain why the palladiumcatalyzed addition of these P(O)OH compounds to alkynes hardly takes place.43

Scheme 19. Competitive Cleavage between P(O)−H Bond and O−H Bond with a Pt(0) Complex

3. SUMMARY In summary, the selective palladium-catalyzed addition of P(O)H compounds to alkynes (hydrophosphorylation) generating alkenylphosphoryl compounds was studied in detail. Four kinds of P(O)H compounds, H-phosphonate (RO)2P(O)H, H-phosphinate R(RO)P(O)H, secondary phosphine oxide R2P(O)H, and hypophosphinic acid H3PO2, can be used as the substrates for the hydrophosphorylation reactions to produce the corresponding Markovnikov adducts in high yields. The use of a right palladium catalyst is the key to a successful hydrophosphorylation, because the best catalyst is different to each other for these P(O)H compounds (Scheme 20). For H-phosphonate (RO)2P(O)H, Pd/dppp enables an efficient hydrophosphorylation. For H-phosphinate R(RO)P(O)H and secondary phosphine oxide R2P(O)H, Pd/dppe/Ph2P(O)OH is the right catalyst. On the other hand, Pd(PPh3)4 is the catalyst for hypophosphinic acid H3PO2. By using this hydrophosphorylation, dozens of grams of alkenylphosphonates can be readily prepared from commercially available starting materials (Scheme 5), showing that this hydrophosphorylation reaction is a powerful practical way for the preparation of these valuable organophosphorus compounds.14,15 This palladium-catalyzed hydrophosphorylation does have limitations. First, it only produces the α-adduct efficiently and selectively.33 Second, except for H3PO2, P(O)H compounds having a free OH group, i.e., phosphinic acids, phosphonic acid, and its monoesters (Figure 2) are not applicable. A few alkynes cannot be used either (Figure 1). Mechanistic studies allow us to draw an overall general catalytic cycle for this palladium-catalyzed hydrophosphorylation (Scheme 21). For (RO)2P(O)H, it reacts like an Brønsted acid to produce the internal palladium intermediate 16 via hydropalladation

Scheme 20. Preferred Catalyst for the Pd-Catalyzed Hydrophosphorylation of Alkynes

ca. 0.5:1 (eq 17 in Scheme 19). Under similar conditions, phosphoric acid (i-PrO)P(O)(OH)H reacted quickly with Pt(PEt3)4 to give 15-4 via activation of an O−H bond in ca. 90% yield, whereas the product 15-4′ via activation of P(O)−H bond was generated in only ca. 10% yield (eq 18 in Scheme 19). When phosphoric acid PhP(O)(OH)H was used, only 15-5 via protonation of Pt(0) was detectable (eq 19 in Scheme 19).

Scheme 21. Mechanism for the Palladium-Catalyzed Addition of P(O)−H Compounds to Alkynes

3152

DOI: 10.1021/jacs.8b00550 J. Am. Chem. Soc. 2018, 140, 3139−3155

Article

Journal of the American Chemical Society

(3) (a) Murphy, P. J. Organophosphorus Reagents; Oxford University Press: Oxford, UK, 2004. (b) Mikolajczyk, M.; Balczewski. Top. Curr. Chem. 2003, 223, 161. (4) (a) Horner, L.; Hoffmann, H.; Wippel, H. G. Chem. Ber. 1958, 91, 61. (b) Bisceglia, J. A.; Orelli, L. R. Curr. Org. Chem. 2015, 19, 744. (c) Tang, W.; Zhang, X. Chem. Rev. 2003, 103, 3029. (5) (a) Baumgartner, T.; Réau, R. Chem. Rev. 2006, 106, 4681. (b) Queffélec, C.; Petit, M.; Janvier, P.; Knight, D. A.; Bujoli, B. Chem. Rev. 2012, 112, 3777. (c) Swanson, J. L. PUREX Process Flowsheets. In Science and Technology of Tributyl Phosphate; Schulz, W. W., Burger, L. L., Navratil, J. D., Bender, K. P., Eds.; CRC Press: Boca Raton, FL, 1984. (d) Fire and Polymers V: Materials and Concepts for Fire Retardancy; Wilkie, C. A., Morgan, A. B., Nelson, G. L., Eds.; American Chemical Society: Washington, DC, 2009; pp 205−248. (e) Green, J. In Fire Retardancy of Polymeric Materials; Grand, A. F., Wilkie, C. A., Eds.; Marcel Dekker: New York, 2000; pp 147−170. (6) (a) Reddy, B. R.; Priya, D. N.; Rao, S. V.; Radhika, P. Hydrometallurgy 2005, 77, 253. (b) Suresh, A.; Srinivasan, T. G.; Rao, P. R. V. Solvent Extr. Ion Exch. 1994, 12, 727. (c) Zhang, Y.; Yu, B.; Wang, B.; Liew, K. M.; Song, L.; Wang, C.; Hu, Y. Ind. Eng. Chem. Res. 2017, 56, 1245. (d) You, G.; Cheng, Z.; Tang, Y.; He, H. Ind. Eng. Chem. Res. 2015, 54, 7309. (7) (a) Schwan, A. L. Chem. Soc. Rev. 2004, 33, 218. (b) Demmer, C. S.; Krogsgaard-Larsen, N.; Bunch, L. Chem. Rev. 2011, 111, 7981. (c) Montchamp, J.-L. Acc. Chem. Res. 2014, 47, 77. (d) Tappe, F. M. J.; Trepohl, V. T.; Oestreich, M. Synthesis 2010, 2010, 3037. (e) Baillie, C.; Xiao, J. Curr. Org. Chem. 2003, 7, 477. (f) Xu, Q.; Han, L.-B. J. Organomet. Chem. 2011, 696, 130. (g) Chen, T.; Zhang, J.-S.; Han, L.-B. Dalton Trans. 2016, 45, 1843. (h) Chen, T.; Han, L.-B. Synlett 2015, 26, 1153. (i) Engel, R.; Cohen, J. I. Synthesis of Carbon-Phosphorus Bonds; CRC Press, New York, 2004. (j) Stockland, R. A. Practical Functional Group Synthesis; John Wiley & Sons: Waukegan, IL, 2016; Chapter 4, pp 219−470. (k) Bange, C. A.; Waterman, R. Chem. - Eur. J. 2016, 22, 12598. (l) Waterman, R. Chem. Soc. Rev. 2013, 42, 5629. (8) Bhattacharya, A. K.; Thyagarajan, G. Chem. Rev. 1981, 81, 415. (9) For radical and acid- or base-catalyzed addition of P(O)−H compounds, see: (a) Organic Phosphorus Compounds; Kosolapoff, G. M., Maier, L., Eds.; Wiley-Interscience: New York, 1972. (b) Goldwhite, H. Introduction to Phosphorus Chemistry; Cambridge University Press: Cambridge, 1981. Selected recent examples: (c) Semenzin, D.; EtemadMoghadam, G.; Albouy, D.; Diallo, O.; Koenig, M. J. Org. Chem. 1997, 62, 2414. (d) Han, L.-B.; Zhao, C.-Q. J. Org. Chem. 2005, 70, 10121. (e) Bunlaksananusorn, T.; Knochel, P. Tetrahedron Lett. 2002, 43, 5817. (f) Mimeau, D.; Gaumont, A. C. J. Org. Chem. 2003, 68, 7016. (g) Hirai, T.; Han, L.-B. Org. Lett. 2007, 9, 53. (h) Isley, N. A.; Linstadt, R. T. H.; Slack, E. D.; Lipshutz, B. H. Dalton Trans. 2014, 43, 13196. (i) Stockland, R. A.; Taylor, R. I.; Thompson, L. E.; Patel, P. B. Org. Lett. 2005, 7, 851. (j) Guo, H.; Yoshimura, A.; Chen, T.; Saga, Y.; Han, L.-B. Green Chem. 2017, 19, 1502. (10) (a) Hirao, T.; Masunaga, T.; Yamada, N.; Ohshiro, Y.; Agawa, T. Bull. Chem. Soc. Jpn. 1982, 55, 909. Hirao’s coupling process has since been studied and modified extensively: (b) Jablonkai, E.; Keglevich, G. Tetrahedron Lett. 2013, 54, 4185. (c) Berger, O.; Petit, C.; Deal, E. L.; Montchamp, J.-L. Adv. Synth. Catal. 2013, 355, 1361. (d) Xu, Y.; Li, Z.; Xia, J.; Guo, H.; Huang, Y. Synthesis 1984, 1984, 781. (e) Kazankova, M. A.; Trostyanskaya, I. G.; Lutsenko, S. V.; Beletskaya, I. P. Tetrahedron Lett. 1999, 40, 569. (f) Stelzer, O.; Sheldrick, W. S.; et al. J. Organomet. Chem. 2002, 645, 14. (g) Pirat, J.-L.; Monbrun, J.; Virieux, D.; Cristau, H.-J. Tetrahedron 2005, 61, 7029. (h) Belabassi, Y.; Alzghari, S.; Montchamp, J.-L. J. Organomet. Chem. 2008, 693, 3171. (i) Deal, E. L.; Petit, C.; Montchamp, J.-L. Org. Lett. 2011, 13, 3270. (j) Rummelt, S. M.; Ranocchiari, M.; van Bokhoven, J. A. Org. Lett. 2012, 14, 2188. (11) (a) Hydrosilylation: A Comprehensive Review on Recent Advances; Marciniec, B., Ed.; Springer: Berlin, 2009. (b) Lipke, M. C.; LibermanMartin, A. L.; Tilley, T. D. Angew. Chem., Int. Ed. 2017, 56, 2260. (c) Nakajima, Y.; Shimada, S. RSC Adv. 2015, 5, 20603. (d) Vogels, C. M.; Westcott, S. A. Curr. Org. Chem. 2005, 9, 687. (e) Trost, B. M.; Ball, Z. T. Synthesis 2005, 6, 853. (f) Xu, L.; Zhang, S.; Li, P. Chem. Soc. Rev. 2015, 44, 8848. (g) Beletskaya, I.; Pelter, A. Tetrahedron 1997, 53, 4957.

which then produces the Markovnikov adduct selectively. On the other hand, for Ph(RO)P(O)H and Ph2P(O)H, they produce a mixture of terminal and internal alkenylpalladium complexes, and consequently give a mixture of the adducts. In the presence of Ph2P(O)OH, hydropalladation of alkynes with this acid takes place first to give an internal alkenylpalladium. A ligand exchange of this complex with a P(O)−H compound gives the internal phosphorylpalladium intermediate which produces the Markovnikov adduct via reductive elimination.



ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/jacs.8b00550. X-ray crystallographic data for 16-6 (CIF) X-ray crystallographic data for 16-8 (CIF) X-ray crystallographic data for 16-9 (CIF) X-ray crystallographic data for 17-6 (CIF) General information, experimental procedures, and characterization data, including 1H, 13C, and 31P NMR spectra for products (PDF)



AUTHOR INFORMATION

Corresponding Author

*[email protected] ORCID

Tieqiao Chen: 0000-0002-9787-9538 Chang-Qiu Zhao: 0000-0002-9016-8151 Li-Biao Han: 0000-0001-5566-9017 Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS Partial financial support from the Canon Foundation, the New Energy and Industrial Technology Development Organization (NEDO) of Japan (Industrial Technology Research Grant Program in 2004), and the Japan Society for the Promotion of Science (JSPS) for the support of Grant-in-Aid for Scientific Research (B) (Kakenhi 16350027) is acknowledged. Thanks are due to Drs. T. Hirai, Y. Ono, and C. Zhang for their help performing some experiments, and Dr. N. Nami for help confirming the structure of complex 12-1.



REFERENCES

(1) (a) Quin, L. D. A Guide to Organophosphorus Chemistry; Wiley Interscience: New York, 2000. (b) Corbridge, D. E. C. Phosphorus: Chemistry, Biochemistry and Technology, 6th ed.; CRC Press: London, 2013. (c) Horsman, G. P.; Zechel, D. L. Chem. Rev. 2017, 117, 5704. (d) Aminophosphonic and Aminophosphinic Acids: Chemistry and Biological Activity; Kukhar, V. P., Hudson, H. R., Eds.; John Wiley & Sons: Chichester, 2000. (e) Ordóñez, M.; Sayago, F. J.; Cativiela, C. Tetrahedron 2012, 68, 6369. (f) Kudzin, Z. H.; Kudzin, M. H.; Drabowicz, J.; Stevens, C. V Curr. Org. Chem. 2011, 15, 2015. (g) Mucha, A.; Kafarski, P.; Berlicki, L. J. Med. Chem. 2011, 54, 5955. (h) Handbook of Organophosphorus Chemistry; Engel, R., Ed.; Marcel Dekker, Inc.: New York, 1992. (i) The Chemistry of Organophosphorus Compounds, Vol. 4; Hartley, F. R., Ed.; John Wiley & Sons: Chichester, 1996. (2) (a) Hendlin, D.; Stapley, E. O.; Jackson, M.; Wallick, H.; Miller, A. K.; Wolf, F. J.; Miller, T. W.; Chaiet, L.; Kahan, F. M.; Foltz, E. L.; Woodruff, H. B.; Mata, J. M.; Hernandez, S.; Mochales, S. Science 1969, 166, 122. (b) John, F. N-Phosphonomethylglycine phytotoxicant compositions. U.S. Patent 3799758, 1974. (c) Hoerlein, G. Rev. Environ. Contam. Toxicol. 1994, 138, 73. 3153

DOI: 10.1021/jacs.8b00550 J. Am. Chem. Soc. 2018, 140, 3139−3155

Article

Journal of the American Chemical Society (h) Burgess, K.; Ohlmeyer, M. J. Chem. Rev. 1991, 91, 1179. (i) Chong, C. C.; Kinjo, R. ACS Catal. 2015, 5, 3238. (j) Takaya, J.; Iwasawa, N. ACS Catal. 2012, 2, 1993. (k) Geier, S. J.; Vogels, C. M.; Westcott, S. A. Current Developments in the Catalyzed Hydroboration Reaction. Boron Reagents in Synthesis; ACS Symposium Series 1236; American Chemical Society: Washington, DC, 2016; Chapter 6, pp 209−225. (12) For coordination of P(O)−H compounds with metals, see: (a) Shaikh, T. M.; Weng, C.-M.; Hong, F.-E Coord. Chem. Rev. 2012, 256, 771. (b) Li, G. Y. Angew. Chem., Int. Ed. 2001, 40, 1513. (c) Wang, X.-B.; Goto, M.; Han, L.-B. Chem. - Eur. J. 2014, 20, 3631. (d) Janesko, B. G.; Fisher, H. C.; Bridle, M. J.; Montchamp, J.-L. J. Org. Chem. 2015, 80, 10025. (e) Duncan, J. A. S.; Hedden, D.; Roundhill, D. M.; Stephenson, T. A.; Walkinshaw, M. D. Angew. Chem., Int. Ed. Engl. 1982, 21, 452. (13) As to how to generally refer to this kind of metal-mediated additions of P(O)−H compounds, we noticed that currently there are several names for them, i.e., hydrophosphorylation, hydrophosphinylation, etc. Since the P(O) group is generally called a phosphoryl group, we prefer to call this kind of reactions hydrophosphorylation. (14) Vinylphosphoryl compounds (taking vinylphosphonates CH2 CHP(O)(OR)2 as an example) are polar monomers similar to acrylates CH2CHCO2R etc. They are readily transformed to other functional compounds through the reactions of the reactive double bonds (addition, reduction, Diels−Alder reaction, epoxidation, dihydroxylation, Heck reaction, metathesis reaction etc.). Therefore, these compounds have vast applications in organic synthesis, medicinal and biochemistry, and material science. For applications in organic synthesis (reviews), see: (a) Minami, T.; Motoyoshiya, J. Synthesis 1992, 1992, 333. (b) Dembitsky, V. M.; Al Quntar, A.; Haj-Yehia, A.; Srebnik, M. Mini-Rev. Org. Chem. 2005, 2, 91. (c) Wang, H.; Liu, Z. Chin. J. Org. Chem. 2003, 23, 321. They are the key intermediates for the preparation of the clinically used antibacterial agent fosfomycin. (d) Girotra, N. N.; Wendler, N. L. Tetrahedron Lett. 1969, 10, 4647. (e) Glamkowski, E. J.; Gal, G.; Purick, R.; Davidson, A. J.; Sletzinger, M. J. Org. Chem. 1970, 35, 3510. For applications in medicinal and biochemistry, a few simple vinylphosphoryl compounds themselves are potential enzyme inhibitors and show strong anti-proliferative activities, see: (f) Al-Quntar, A. A. A.; Baum, O.; Reich, R.; Srebnik, M. Arch. Pharm. 2004, 337, 76. (g) Wang, X.-H.; Han, L.-B. Jpn. Kokai Tokkyo Koho JP 200763168, 2007. (h) Wang, X.-H.; Han, L.-B. Jpn. Kokai Tokkyo Koho JP 2007137838, 2007. (i) Wang, X.-H.; Han, L.-B. Jpn. Kokai Tokkyo Koho JP 200770278, 2007. Examples for biological applications of their derivatives as anti-cancer, anti-viral, and anti-bacterial compounds, see:. (j) Liu, Z.; MacRitchie, N.; Pyne, S.; Pyne, N. J.; Bittman, R. Bioorg. Med. Chem. 2013, 21, 2503. (k) Tonelli, F.; Lim, K. G.; Loveridge, C.; Long, J.; Pitson, S. M.; Tigyi, G.; Bittman, R.; Pyne, S.; Pyne, N. J. Cell. Signalling 2010, 22, 1536. (l) Harnden, M. R.; Parkin, A.; Parratt, M. J.; Perkins, R. M. J. Med. Chem. 1993, 36, 1343. (m) Lazrek, H. B.; Rochdi, A.; Khaider, H.; Barascut, J. L.; Imbach, J. L.; Balzarini, J.; Witvrouw, M.; Pannecouque, C.; De Clercq, E. Tetrahedron 1998, 54, 3807. Because they can homo- or copolymerize to give functional polymers, huge applications in material chemistry are known, i.e., flame retardant materials, metal extractants, fuel cell membranes, thermoresponsive polymers, cement, dental materials, etc. (n) David, G.; Negrell-Guirao, C. In Phosphorus-Based Polymers: from Synthesis to Applications; Monge, S., David, G., Eds.; RSC Polymer Chemistry Series 11; Royal Society of Chemistry: United Kingdom, 2014; pp 35−50. (o) Soller, B. S.; Salzinger, S.; Rieger, B. Chem. Rev. 2016, 116, 1993. See also: (p) Han, L.-B.; Uchimaru, Y.; Sasaki, S. Jpn. Kokai Tokkyo Koho JP 2013253816, 2013. (q) Han, L.-B.; Liu, R. Jpn. Kokai Tokkyo Koho JP 2010179287, 2010. (r) Han, L.-B.; Futamura, Y.; Fujino, H.; Watanabe, T. Jpn. Kokai Tokkyo Koho JP 2015110617, 2015. (s) Han, L.-B.; Futamura, Y.; Fujino, H.; Watanabe, T. Jpn. Kokai Tokkyo Koho JP 2014132089, 2014. (t) Han, L.-B.; Shinohara, Y. Jpn. Kokai Tokkyo Koho JP 2007145951, 2007. Recently, a remarkable application of these vinylphosphoryl compounds that can significantly improve the stability and performance of lithium battery even at high temperatures was discovered: (u) Han, J.; Lee, H.; Kim, H. U.S. Patent US 7217480B2, 2007. (v) Han, L.-B; Matumoto, H.; Yoshinaga, M.; Yamashita, H. Jpn.

Kokai Tokkyo Koho JP 2014205637, 2014. (w) Funada, Y.; Kubota, T. Jpn. Kokai Tokkyo Koho JP 201355031, 2013. (15) A few methods for the preparation of vinylphosphoryl compounds are known: (A) The early dehydrobromination method of 2-bromoethylphosphonate which was prepared from (EtO)3P and BrCH2CH2Br under high temperatures: (a) Kosolapoff, G. M. J. Am. Chem. Soc. 1948, 70, 1971. (B) Metal-catalyzed couplings of halides (ref 10). Recently, C−O, C−S, C−CN, etc. bonds have also been converted to C−P bonds via this coupling method: (b) Yang, J.; Chen, T.; Han, L.B. J. Am. Chem. Soc. 2015, 137, 1782. (c) Yang, J.; Xiao, J.; Chen, T.; Han, L.-B. J. Org. Chem. 2016, 81, 3911. (d) Yang, J.; Xiao, J.; Chen, T.; Yin, S.-F.; Han, L.-B. Chem. Commun. 2016, 52, 12233. (e) Zhang, J.-S.; Chen, T.; Yang, J.; Han, L.-B. Chem. Commun. 2015, 51, 7540. (C) Other methods such as the transformation of alkynylphosphonates etc.: (f) Mulla, K.; Aleshire, K. L.; Forster, P. M.; Kang, J. Y. J. Org. Chem. 2016, 81, 77 and references cited therein. (16) (a) Han, L.-B.; Tanaka, M. J. Am. Chem. Soc. 1996, 118, 1571. It seems that there are two early examples of metal-mediated P(O)−H additions to carbon−carbon unsaturated bonds before this work. First, a U.S. patent claimed the addition of (RO)2P(O)H to alkynes by Ni(PR3)2X2 at elevated temperatures. See: (b) Lin, K. U.S. Patent 3681481, 1972. Hirao and co-workers also very briefly claimed in their paper, describing the coupling of (RO)2P(O)H with organic halides, that (RO)2P(O)H added to isoprene, though the efficiency was low (10% yield, 150 °C, 20 h). See ref 10a. (17) (a) Delacroix, O.; Gaumont, A. C. Curr. Org. Chem. 2005, 9, 1851. (b) Wicht, D. K.; Glueck, D. S. Hydrophosphination and related reactions. In Catalytic heterofunctionalization; Togni, A., Grützmacher, H., Eds.; Wiley-VCH: Weinheim, 2001; Chapter 5, pp 143−170. (c) Greenberg, S.; Stephan, D. W. Chem. Soc. Rev. 2008, 37, 1482. (d) Beletskaya, I. P.; Kabachnik, M. M. Mendeleev Commun. 2008, 18, 113. (e) Coudray, L.; Montchamp, J.-L. Eur. J. Org. Chem. 2008, 2008, 3601. (18) (a) http://www.katayamakagaku.co.jp/products/ chemicalproducts/flameretardant/index.html. (b) Henkelmann, J.; Klass, K.; Arndt, J. Eur. Patent 1203773, 2002. http://www.basf.de/ en/intermed/products/vinylphosphonicacid/. (c) http://www. euticals.com/attachments/article/11/Euticals_VPA_final.pdf. (19) (a) Han, L.-B.; Hua, R.; Tanaka, M. Angew. Chem., Int. Ed. 1998, 37, 94. (b) Han, L.-B.; Zhao, C.-Q.; Onozawa, S.-y.; Goto, M.; Tanaka, M. J. Am. Chem. Soc. 2002, 124, 3842. (c) Han, L.-B.; Zhang, C.; Yazawa, H.; Shimada, S. J. Am. Chem. Soc. 2004, 126, 5080. (d) Han, L.-B.; Ono, Y.; Yazawa, H. Org. Lett. 2005, 7, 2909. (e) Han, L.-B.; Choi, N.; Tanaka, M. Organometallics 1996, 15, 3259. (f) Kanada, J.; Tanaka, M. Adv. Synth. Catal. 2011, 353, 890. (g) Xu, Q.; Shen, R.; Ono, Y.; Nagahata, R.; Shimada, S.; Goto, M.; Han, L.-B. Chem. Commun. 2011, 47, 2333. (20) Part of this work was reported in 225th National Meeting of the American Chemical Society: Han, L.-B. Abstracts of Papers of the American Chemical Society, New Orleans, Louisiana, Mar 23−27, 2003; paper no. U148. (21) Stockland et al. have investigated the reductive elimination step for the formation of the C−P(O) bonds using C−Pd−P(O) complexes. (a) Kohler, M. C.; Grimes, T. V.; Wang, X.; Cundari, T. R.; Stockland, R. A. Organometallics 2009, 28, 1193. (b) Kohler, M. C.; Stockland, R. A.; Rath, N. P. Organometallics 2006, 25, 5746. (c) Stockland, R. A.; Levine, A. M.; Giovine, M. T.; Guzei, I. A.; Cannistra, J. C. Organometallics 2004, 23, 647. (d) Levine, A. M.; Stockland, R. A.; Clark, R.; Guzei, I. Organometallics 2002, 21, 3278. The reaction has also been studied by calculation etc.: (e) Wang, T.; Sang, S.; Liu, L.; Qiao, H.; Gao, Y.; Zhao, Y. J. Org. Chem. 2014, 79, 608. (f) Ananikov, V. P.; Khemchyan, L. L.; Beletskaya, I. P.; Starikova, Z. A. Adv. Synth. Catal. 2010, 352, 2979. For the reductive elimination of Ar−Pd−P(O)(OEt)2, an electron-poor aryl reacted slower than that of an electron-rich aryl group (ref 21a). (22) Previous communications related to cat Pd/(RO)2P(O)H/ alkyne generating alkenylphosphonates: (i) First communicated with the Me2Pd(PPh2Me)2-mediated addition of (RO)2P(O)H (R = Me, Et) (ref 16a). (ii) An oxapalladacycle-mediated addition to 1-octyne and tert-butylacetylene (ref 19g). (iii) Reactions with α,ω-diynes generating 3154

DOI: 10.1021/jacs.8b00550 J. Am. Chem. Soc. 2018, 140, 3139−3155

Article

Journal of the American Chemical Society

cated using a Pd/xantphos catalyst with 1-octyne, phenylacetylene, and 4-octyne: see: (a) Deprele, S.; Montchamp, J.-L. J. Am. Chem. Soc. 2002, 124, 9386. (b) Deprele, S.; Montchamp, J.-L. Org. Lett. 2004, 6, 3805. (35) A separate experiment confirmed that the alkenylphosphoryl compounds can be reduced to the saturated 9. H3PO2 is an efficient reductant for transition metal-catalyzed transfer hydrogenation of unsaturated carbon-carbon bonds. See: (a) Guyon, C.; Metay, E.; Popowycz, F.; Lemaire, M. Org. Biomol. Chem. 2015, 13, 7879 and references therein. (b) Oliveira, M. C. F. Appl. Catal., A 2007, 329, 7. (c) Kawasaki, T.; Tanaka, H.; Tsutsumi, T.; Kasahara, T.; Sato, I.; Soai, K. J. Am. Chem. Soc. 2006, 128, 6032. (36) In contrast to trans-[(MeO)2P(O)]2Pd(PEt3)2, P−P couplings could not be clearly observed for trans-[(MeO)2P(O)]2Pd(PPh2Me)2, perhaps due to the weak ligating ability of PPh2Me that may dissociate from the complex in the solution. The combustion elemental analysis agrees with its composition. (37) As confirmed separately, the possible isomerization between the cis- and trans-Me2Pd(PEt3)2 in C6D6 is negligible under current conditions. (38) Under similar conditions, in the absence of (MeO)2P(O)H, cisMe2Pd(PEt3)2 is stable with water. (39) The small germinal coupling between the two vinyl protons was not recognizable in the presence of a free PEt3 due to rapid exchange reactions between the complex and PEt3. (40) For hydropalladation of carboxylic acids with the combination of Pd(PEt3)4 and alkynes, see: (a) Shen, R.; Chen, T.; Zhao, Y.; Qiu, R.; Zhou, Y.; Yin, S.; Wang, X.; Goto, M.; Han, L.-B. J. Am. Chem. Soc. 2011, 133, 17037. We noted that a protonation mechanism was proposed for the hydropalladation of Pd-alkyne complex with Ph2P(O)OH on the basis of DFT studies. See: (b) Zhang, H.; Bao, X. RSC Adv. 2015, 5, 84636. Although we do not have evidences excluding path b, its contribution should be small since it can hardly explain the high yields of the alkenylpalladium complexes via a long-time slow reaction, considering the rapid decomposition of the H−Pd complex (Scheme 14). Furthermore, the regioselectivity of the hydropalladation is also similar to the reactions of carboxylic acids. Unfortunately, details on the insertion of an alkyne were not clear. We thank referees who suggested alternative mechanisms such as a nucleophilic addition of a hydride path. In addition, the alkyne path (path a) and the results of Scheme 7 would not be mutually exclusive considering an electron-poor alkyne coordinates easier with an alkyne. (41) In a real sense, the results shown in Table 11 may only reflect the stability of complex 16, because a reductive elimination required the isomerization of trans-16 to cis-16 or a dissociation of PEt3 occurred first, that, unfortunately, was not available at this stage. (42) Li, J.; Liu, L.; Fu, Y.; Guo, Q. Tetrahedron 2006, 62, 4453. (43) This study was first conducted using Pd(PEt3)4. However, complicated results were obtained due to the rapid decomposition of the unstable Pd−H species.

cyclopentenes: (a) Kanada, J.; Yamashita, K.; Nune, S.; Tanaka, M. Tetrahedron Lett. 2009, 50, 6196. (23) For the preparation and properties of these palladium complexes, see: (a) Ozawa, F.; Ito, T.; Nakamura, Y.; Yamamoto, A. Bull. Chem. Soc. Jpn. 1981, 54, 1868. (b) De Graaf, W.; Boersma, J.; Smeets, W. J. J.; Spek, A. L.; Van Koten, G. Organometallics 1989, 8, 2907. (24) The starting materials (1-octyne, (MeO)2P(O)H, and toluene) used in this reaction were all dried and freshly distilled. (25) 1-Octyne, phenylacetylene, and (MeO)2P(O)H were purchased from TCI. Caution! For safety concerns, the use of the catalyst Pd(OAc)2/ dppp without solvent should be avoided with the reaction of an aromatic alkyne because the author (L.-B.H.) had observed an explosion when the reaction of phenylacetylene was carried out without solvent using Pd(OAc)2/ dppp as the catalyst. The reason was not clear. It was assumed that an explosive metal phenylacetylide was generated via the reaction of phenylacetylene with Pd(OAc)2. (26) Han, L.-B.; Ono, Y.; Xu, Q.; Shimada, S. Bull. Chem. Soc. Jpn. 2010, 83, 1086. (27) This five-membered H-phosphonate was known to be more reactive than other (RO)2P(O)H compounds. For discussions, see: Han, L.-B.; Mirzaei, F.; Zhao, C.-Q.; Tanaka, M. J. Am. Chem. Soc. 2000, 122, 5407 and references cited therein. (28) For Pd-catalyzed head-to-tail dimerization forming enynes, see: (a) Chen, T.; Guo, C.; Goto, M.; Han, L.-B. Chem. Commun. 2013, 49, 7498. (b) Trost, B. M.; Chan, C.; Ruhter, G. J. Am. Chem. Soc. 1987, 109, 3486. (c) Trost, B. M.; Sorum, M. T.; Chan, C.; Ruhter, G. J. Am. Chem. Soc. 1997, 119, 698. (29) We do not have evidence concerning the mechanism. However, as shown below, the terminal diyne may undergo a similar intermolecular head-to-tail cyclization (see Scheme 8), generating an enyne 7, which further reacts with (MeO)2P(O)H to give 8.

Isomerization of 8 by palladium would generate 6. For a Pdcatalyzed cyclization of aliphatic terminal diynes via intermolecular head-to-tail dimerization forming cyclic enynes, see: Lucking, U.; Pfaltz, A. Synlett 2000, 9, 1261.

(30) Because tert-butylacetylene also gives the α-adduct as the major product (eq 6), this selectivity could not be simply due to steric reasons. The well-known feature that a silyl group stabilizes an α-anion and βcation might contribute to this selectivity. (31) Previous communications related to cat Pd/R′ (RO)P(O)H/ alkyne generating alkenylphosphinates: (i) First communicated by the Me2Pd(PPhMe2)2-mediated addition of Ph(MenO)P(O)H (Men = menthyl) (ref 19b). (ii) An oxapalladacycle-mediated addition of Ph(EtO)P(O)H to 1-octyne and tert-butylacetylene (ref 19g). (iii) Pd(OAc)2/phosphine-mediated additions: (a) Kumar Nune, S.; Tanaka, M. Chem. Commun. 2007, 50, 2858. (32) Previous communications related to cat Pd/RR′P(O)H/alkyne generating alkenylphosphine oxides: (i) First communicated with a Pd(PPh 3 ) 4 -mediated addition of Ph 2 P(O)H (ref 19e). (ii) Me2Pd(PPhMe2)/Ph2P(O)OH-mediated addition (ref 19a). (iii) An oxapalladacycle-mediated addition to 1-octyne, tert-butylacetylene, and diphenylacetylene (ref 19g). (iii) Pd(OAc)2/phosphine-mediated additions of Ph2P(O)H etc: Dobashi, N.; Fuse, K.; Hoshino, T.; Kanada, J.; Kashiwabara, T.; Kobata, C.; Tanaka, M. Tetrahedron Lett. 2007, 48, 4669. (33) Although Pd(PPh3)4 could catalyze the addition of Ph2P(O)H with terminal alkynes to give the β-adducts as the major products, these compounds were more conveniently prepared by the Rh-catalyzed P(O)−H additions which almost perfectly produce the β-adducts (ref 7f). A Cu-catalyzed addition also produces the β-adducts: Niu, M.; Fu, H.; Jiang, Y.; Zhao, Y. Chem. Commun. 2007, 272. (34) Previous communications related to cat Pd/H2P(O)(OH)/ alkyne generating hydrogen alkenylphosphinic acids: First communi3155

DOI: 10.1021/jacs.8b00550 J. Am. Chem. Soc. 2018, 140, 3139−3155